During 2005-2006, research in AMRI was focused on 2 main projects: 1) the investigation of brain activity patterns during rest and sleep; and 2) the development of MRI techniques on NIH's new 7.0 T human scanner. [unreadable] [unreadable] The investigation of brain activity fluctuations during rest was a continuation of earlier research in AMRI. In the previous years, we found that during stage 1 and 2 sleep, metabolic fluctuations occur in distinct brain regions with an apparent functional relationship. We hypothesized that these fluctuations support synaptic pruning and strengthening to consolidate traces from waking activity. Previous indications of this activity arose during EEG experiments performed by the Tononi Sleep Laboratory in Madison. To test this hypothesis, a collaboration was started with Allen Braun (NIDCS) and Tom Balkin (WRAIR) to study the entire sleep-wake cycle with simultaneous EEG/fMRI. So far, about 20 subjects have been studied and preliminary data analysis has started.[unreadable] [unreadable] During 2005-2006, the development of the 7.0 T human scanner involved 4 projects. First, a method was developed to perform high resolution MRI based on T2* magnitude and phase contrast, and it was found that a substantial contrast between and within grey and white matter exists. This contrast allows in-vivo cortical imaging with unprecedented spatial resolution. A potential important contributor to this contrast is tissue iron, which has been implicated as an indicator of a variety of disease processes. Currently, the origin of this contrast is investigated in both human tissue samples (from NCI) and animal models (in collaboration with Afonso Silva). Second, in collaboration with Pjotr Starewicz from Resonance Research, a method was developed and tested to image quality by correct to magnetic field changes related to the respiratory cycle. This method involves adjusting the magnetic field shims in real-time. An EIR has been filed for this method. Third, in collaboration with Helmut Merkle and Shumin Wang, we started a project aimed at improving B1 uniformity in human brain to improve image quality using a novel strategy. Preliminary data show substantial uniformity improvement. Also for this method, an EIR was filed. Fourth, a project was started to correct for human head motion in real-time. Head motion is one of the main limiting factors for resolution improvement for in-vivo MRI. To allow real-time control of the MRI scanner, a collaboration was started with Stanford University.